DESCRIPTION (Applicant's abstract): Depression among primary care patients has been shown to be underrecognized and undertreated. However, previous research has not determined which physicians do not reliably diagnose or treat depression and why they make such errors. Thus, it is critical to address the main objective of this proposal: to determine patient and physician factors that are associated with physicians' quality of care for depression. The specific aims for this project are: 1) To describe primary care physicians' decision-making for depression care in the medically ill; 2) To determine the relationship of patient variables (medical illness comorbidity, attributional style, attitudes toward depression treatment, and demographics) with physicians' decisions for the care of depression; and 3) To determine the relationship of physician characteristics (e.g. specialty, experience with patients with depression, age, and gender) with physicians' decisions for the care of depression. The overall design of the study is the presentation of one videotaped case of a patient with major depression and physical complaints to each physician subject. The study has a 2x2x2x2x2 factorial design with two main effects: medical illness comorbidity (recent myocardial infarction vs. good health); and attributional style (somatic vs. non-somatic). The vignettes will also vary by treatment preference (prefer no mental health treatment vs. accepting of treatment), race (African-American vs. White), and gender. Thus, there will be 32 vignettes, each of which represents one distinct combination of levels. Subjects will be 500 physicians, half general internists and half family physicians. Each physician will complete a semi-structured interview assessing diagnostic, evaluation and treatment decisions for the case vignette. This interview will be followed by a structured interview assessing covariates including physician experience with patients with depression, demographic variables, practice characteristics, and attitudes regarding the care of patients with depression. The direct benefit of this study would be enhanced understanding of patient and physician characteristics that determine whether depression is appropriately treated. The results from this study can guide continuing medical education efforts and be used to improve future intervention studies designed to enhance quality of care.